Solution NMR of signal peptidase, a membrane protein

Useful solution nuclear magnetic resonance (NMR) data can be obtained from full-length, enzymatically active type I signal peptidase (SPase I), an integral membrane protein, in detergent micelles. Signal peptidase has two transmembrane segments, a short cytoplasmic loop, and a 27-kD C-terminal catalytic domain. It is a critical component of protein transport systems, recognizing and cleaving amino-terminal signal peptides from preproteins during the final stage of their export. Its structure and interactions with the substrate are of considerable interest, but no three-dimensional structure of the whole protein has been reported. The structural analysis of intact membrane proteins has been challenging and only recently has significant progress been achieved using NMR to determine membrane protein structure. Here we employ NMR spectroscopy to study the structure of the full-length SPase I in dodecylphosphocholine detergent micelles. HSQC-TROSY spectra showed resonances corresponding to approximately 3/4 of the 324 residues in the protein. Some sequential assignments were obtained from the 3D HNCACB, 3D HNCA, and 3D HN(CO) TROSY spectra of uniformly 2H, 13C, 15N-labeled full-length SPase I. The assigned residues suggest that the observed spectrum is dominated by resonances arising from extramembraneous portions of the protein and that the transmembrane domain is largely absent from the spectra. Our work elucidates some of the challenges of solution NMR of large membrane proteins in detergent micelles as well as the future promise of these kinds of studies.     

Solution NMR of signal peptidase, a membrane protein

Useful solution nuclear magnetic resonance (NMR) data can be obtained from full-length, enzymatically active type I signal peptidase (SPase I), an integral membrane protein, in detergent micelles. Signal peptidase has two transmembrane segments, a short cytoplasmic loop, and a 27-kD C-terminal catalytic domain. It is a critical component of protein transport systems, recognizing and cleaving amino-terminal signal peptides from preproteins during the final stage of their export. Its structure and interactions with the substrate are of considerable interest, but no three-dimensional structure of the whole protein has been reported. The structural analysis of intact membrane proteins has been challenging and only recently has significant progress been achieved using NMR to determine membrane protein structure. Here we employ NMR spectroscopy to study the structure of the full-length SPase I in dodecylphosphocholine detergent micelles. HSQC-TROSY spectra showed resonances corresponding to approximately 3/4 of the 324 residues in the protein. Some sequential assignments were obtained from the 3D HNCACB, 3D HNCA, and 3D HN(CO) TROSY spectra of uniformly ²H, ¹³C, ¹⁵N-labeled full-length SPase I. The assigned residues suggest that the observed spectrum is dominated by resonances arising from extramembraneous portions of the protein and that the transmembrane domain is largely absent from the spectra. Our work elucidates some of the challenges of solution NMR of large membrane proteins in detergent micelles as well as the future promise of these kinds of studies.     

The microvillus 110K cytoskeletal protein is an integral membrane protein

A protein of 110,000 MW connects actin filaments to the plasma membrane in microvilli of intestinal epithelial cells. In the present study four independent lines of evidence suggest that the 110K protein is directly bound to the lipid bilayer. The solubilization of the 110K protein requires detergents and removal of detergent after solubilization results in aggregation. The 110K protein partitions into the detergent phase in Triton X-114 solutions. It is selectively incorporated into liposomes. It is specifically labeled with the hydrophobic probe ¹⁴C-phenylisothiocyanate. In addition we present a purification scheme for the 110K protein in milligram amounts. This represents the simplest system of membrane to filament attachment, in which an integral membrane protein is also a cytoskeletal protein.     

Purification and characterization of hen oviduct microsomal signal peptidase

Hen oviduct signal peptidase requires only two proteins for proteolysis of fully synthesized secretory precursor proteins in vitro: one with a molecular mass of 19 kilodaltons (kDa) and one which is a glycoprotein whose mass varies from 22 to 24 kDa depending on the extent of glycosylation. Purified signal peptidase has been analyzed both as part of an active catalytic unit and after electroelution of the individual proteins out of a preparative polyacrylamide gel. The multiple forms of the glycoprotein component of signal peptidase bind to concanavalin A and are shown to be derived from the same polypeptide backbone. Removal of their oligosaccharides by digestion with N-glycanase converts these proteins to a single 19.5-kDa polypeptide. The glycoproteins all exhibit very similar profiles following individual digestion with trypsin and separation of the resulting peptides by reverse-phase high-performance liquid chromatography. In addition, sequence analysis of selected peptides from corresponding regions in chromatograms representing each form of the glycoprotein reveals the same amino acid sequences. The 19-kDa signal peptidase protein does not bind concanavalin A, has a distinct tryptic peptide map from that of the glycoprotein, and appears to share no amino acid sequences in common with the glycoprotein. Its copurification on a concanavalin A-Sepharose column indicates that it must interact directly with the glycoprotein subunit.     

Partial purification of microsomal signal peptidase from hen oviduct

Signal peptidase has been purified approximately 600-fold from hen oviduct microsomes. Treatment of microsomes with ice-cold sodium carbonate at pH 11.5 removes soluble and extrinsic membrane proteins prior to solubilization of signal peptidase with Nonidet P-40. After dialysis to pH 8.2, the solubilized enzyme is chromatographed on diethylaminoethyl cellulose at pH 8.2. More than 90% of contaminating proteins bind to the column while signal peptidase and endogenous phospholipid are eluted in the column void volume. Enzyme activity subsequently binds to carboxymethyl cellulose at pH 5.8 and is eluted by approximately 100 to 200 mM NaCl during a NaCl gradient. Polypeptides present in partially purified hen oviduct signal peptidase have relative molecular masses ranging from 54 kD to less than 11 kD with major bands at 29, 23, 22, 19, 18 and 13 kD. The purified peptidase requires phospholipid for activity and is maximally active in the presence of 2 mg/ml phosphatidylcholine.     

Evidence that eel acetylcholinesterase is not an integral membrane protein

Detergent binding studies indicated that the neural enzyme, acetylcholinesterase, did not exhibit the properties of an integral membrane protein. The 11S form was isolated by affinity chromatography from a tryptic digest and the 14S and 18S forms in like manner from an undigested preparation. Studies were performed with [³H]TX-100 to determine the extent of binding by these forms and with catalase and human low density lipoprotein as reference proteins. All forms of the enzyme bound less than 0.04 mg TX-100/mg protein which is only slightly higher than binding by catalase and about 25 fold lower than the binding exhibited by low density lipoprotein.     

Chromogranin, an integral membrane protein

Chromogranin is the major soluble protein of the adrenal medulla chromaffin granule and is secreted upon nervous stimulation. Using antisera to pure chromogranin in immunoblotting procedures, we show that chromogranin is the major integral membrane protein as well. Extraction of chromaffin granule membranes with low salt, high salt, chelating agents, or calcium-containing solutions does not remove the chromogranin from the membranes. The membrane form of chromogranin can be purified on a C-18 semi-preparative column using high pressure liquid chromatography. Amino-terminal sequence data indicate that the membrane and soluble forms of chromogranin are identical or very similar.     

Electrostatic interactions in an integral membrane protein

The effects of charge-charge interactions on the midpoint reduction potential (E(m)()) of the primary electron donor (P) in the photosynthetic reaction center of Rhodobacter sphaeroides were investigated by introducing mutations of ionizable amino acids at selected sites. The mutations were designed to alter the electrostatic environment of P, a bacteriochlorophyll dimer, without greatly affecting its structure or molecular orbitals. Two arginine residues at homologous positions in the L and M subunits [residues (L135) and (M164)], Asp (L155), Tyr (L164), and Cys (L247) were changed independently. Arginine (L135) was replaced by Lys, Leu, Gln, or Glu; Arg (M164), by Leu or Glu; Asp (L155), by Asn; Tyr (L164), by Phe; and Cys (L247), by Lys or Asp. The R(L135)E/C(L247)K double mutant also was made. The shift in the E(m)() of P/P(+) was measured in each mutant and was compared with the effect predicted by electrostatics calculations using several different computational approaches. A simple distance-dependent dielectric screening factor reproduced the effects remarkably well. By contrast, microscopic methods that considered the reaction field in the protein and solvent but did not include explicit counterions overestimated the changes in the E(m)() considerably. Including counterions for the charged residues reduced the calculated effects of the mutations in molecular dynamics calculations. The results show that electrostatic interactions of P with ionizable amino acid residues are strongly screened, and suggest that counterions make major contributions to this screening. The screening also could reflect penetration of water or other relaxations not taken into account because of incomplete sampling of configurational space.     

Hen oviduct N alpha-acetyltransferase is a ribonucleoprotein having 7 S RNA

Hen oviduct N alpha-acetyltransferase was clarified to have a nucleic acid as an existing constituent by the following three results: (i) an ultraviolet absorption spectrum of the purified N alpha-acetyltransferase free of S-acetyl coenzyme A (Ac-CoA) had an absorption maximum at 260 nm. (ii) A nucleic acid band stained with ethidium bromide was detected on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. (iii) An ethidium bromide band co-migrated with a fluorescent band of the protein treated with N-(7-dimethylamino-4-methylcoumarinyl)maleimide, a reagent specific for thiol groups, on polyacrylamide gel electrophoresis in the absence of sodium dodecyl sulfate. N alpha-Acetyltransferase lost its activity partially or completely by digestion with bovine pancreatic RNase A, Staphylococcus aureus nuclease, or proteinase K, showing that both the nucleic acid and the protein subunit were necessary for the enzyme activity. The nucleic acid component was identified as an RNA but not a DNA because the RNase T2 digest of the nucleic acid was composed of four 3'-ribomononucleotides and completely separated from 3'- and 5'-deoxyribomononucleotides on TLC. The chain length of the nucleic acid of 260 nucleotides estimated by formamide-polyacrylamide gel electrophoresis was calculated to be about 83,000 of the molecular weight. The contents of RNA (35.0%) and protein (65.0%) in N alpha-acetyltransferase determined on weight basis corresponded reasonably well to the contents of RNA (34.4%) and protein (65.6%) calculated based on the assumption that N alpha-acetyltransferase consisted of one molecule of 7 S RNA (Mr 83,000) and two identical Mr 79,000 protein subunits. The total molecular weight (241,000) of the holoenzyme calculated based on the above result was identical to the molecular weight (240,000) of N alpha-acetyltransferase estimated by Sepharose 6B gel filtration.     

Drosophila signal peptide peptidase is an essential protease for larval development

We identified the Drosophila melanogaster Signal peptide peptidase gene (Spp) that encodes a multipass transmembrane aspartyl protease. Drosophila SPP is homologous to the human signal peptide peptidase (SPP) and is distantly related to the presenilins. We show that, like human SPP, Drosophila SPP can proteolyze a model signal peptide and is sensitive to an SPP protease inhibitor and that it localizes to the endoplasmic reticulum. Expression of Drosophila SPP was first apparent at germ band extension, and in late embryos it was robust in the salivary glands, proventriculus, and tracheae. Flies bearing mutations in conserved residues or carrying deficiencies for the Spp gene had defective tracheae and died as larvae.     

Conformationally specific misfolding of an integral membrane protein

Membrane protein misfolding is related to the etiology of many diseases, but is poorly understood, particularly from a structural standpoint. This study focuses upon misfolding of a mutant form of diacylglycerol kinase (s-DAGK), a 40 kDa homotrimeric protein having nine transmembrane segments. Preparations of s-DAGK sometimes contain a kinetically trapped misfolded population, as evidenced by lower-than-expected enzyme activity (with no accompanying change in substrate K(m)) and by the appearance of a second band in electrophoresis gels. Misfolding of s-DAGK may take place during cellular overexpression, but can also be reproduced using the purified enzyme. TROSY NMR spectra of s-DAGK as a 100 kDa complex with detergent micelles exhibit a single additional set of resonances from the misfolded form, indicating a single misfolded conformational state. The relative intensities of these extra resonances correlate with the percent reduction in enzyme activity below the maximum observed for fully folded s-DAGK. Misfolded s-DAGK exhibits a modest difference in its far-UV CD spectrum compared to the folded enzyme, consistent with a small degree of variance in secondary structural content between the two forms. However, differences in NMR chemical shift dispersion and temperature-dependent line widths exhibited by folded and misfolded s-DAGK support the notion that they represent very different structural states. Cross-linking experiments indicate that both the correctly folded enzyme and the kinetically trapped misfolded form are homotrimers. This work appears to represent the first documentation of conformationally specific misfolding of an integral membrane protein.     

Human immunodeficiency virus type 1 Vpu protein is an oligomeric type I integral membrane protein.

The human immunodeficiency virus type 1 Vpu protein is a 16-kDa phosphoprotein which enhances the efficiency of virion production and induces rapid degradation of CD4, the cellular receptor for human immunodeficiency virus. The topology of membrane-inserted Vpu was investigated by using in vitro-synthesized Vpu cotranslationally inserted into canine microsomal membranes. Proteolytic digestion and immunoprecipitation studies revealed that Vpu was a type I integral membrane protein, with the hydrophilic domain projecting from the cytoplasmic membrane face. In addition, several high-molecular-weight proteins containing Vpu were identified by chemical cross-linking. Such complexes also formed when wild-type Vpu and a Tat-Vpu fusion protein were coexpressed. Subsequent analysis by one- and two-dimensional electrophoresis revealed that these high-molecular-weight complexes consisted of homo-oligomers of Vpu. These findings indicate that Vpu is a type I integral membrane protein capable of multimerization.     

The chloroplast import receptor is an integral membrane protein of chloroplast envelope contact sites

A chloroplast import receptor from pea, previously identified by antiidiotypic antibodies was purified and its primary structure deduced from its cDNA sequence. The protein is a 36-kD integral membrane protein (p36) with eight potential transmembrane segments. Fab prepared from monospecific anti-p36 IgG inhibits the import of the ribulose-1,5- bisphosphate carboxylase small subunit precursor (pS) by interfering with pS binding at the chloroplast surface. Anti-p36 IgGs are able to immunoprecipitate a Triton X-100 soluble p36-pS complex, suggesting a direct interaction between p36 and pS. This immunoprecipitation was specific as it was abolished by a pS synthetic transit peptide, consistent with the transit sequence receptor function of p36. Immunoelectron microscopy localized p36 to regions of the outer chloroplast membrane that are in close contact with the inner chloroplast membrane. Comparison of the deduced sequence of pea p36 to that of other known proteins indicates a striking homology to a protein from spinach chloroplasts that was previously suggested to be the triose phosphate-3-phosphoglycerate-phosphate translocator (phosphate translocator) (Flugge, U. I., K. Fischer, A. Gross, W. Sebald, F. Lottspeich, and C. Eckerskorn. 1989. EMBO (Eur. Mol. Biol. Organ.) J. 8:39-46). However, incubation of Triton X-100 solubilized chloroplast envelope material with hydroxylapatite indicated that p36 was quantitatively absorbed, whereas previous reports have shown that phosphate translocator activity does not bind to hydroxylapatite (Flugge, U. I., and H. W. Heldt. 1981. Biochim. Biophys. Acta. 638:296- 304. These data, in addition to the topology and import inhibition data presented in this report support the assignment of p36 as a receptor for chloroplast protein import, and argue against the assignment of the spinach homologue of this protein as the chloroplast phosphate translocator.     

AtCSLD2 is an integral Golgi membrane protein with its N-terminus facing the cytosol

Cellulose synthase-like proteins in the D family share high levels of sequence identity with the cellulose synthase proteins and also contain the processive beta-glycosyltransferase motifs conserved among all members of the cellulose synthase superfamily. Consequently, it has been hypothesized that members of the D family function as either cellulose synthases or glycan synthases involved in the formation of matrix polysaccharides. As a prelude to understanding the function of proteins in the D family, we sought to determine where they are located in the cell. A polyclonal antibody against a peptide located at the N-terminus of the Arabidopsis D2 cellulose synthase-like protein was generated and purified. After resolving Golgi vesicles from plasma membranes using endomembrane purification techniques including two-phase partitioning and sucrose density gradient centrifugation, we used antibodies against known proteins and marker enzyme assays to characterize the various membrane preparations. The Arabidopsis cellulose synthase-like D2 protein was found mostly in a fraction that was enriched with Golgi membranes. In addition, versions of the Arabidopsis cellulose synthase-like D2 proteins tagged with a green fluorescent protein was observed to co-localize with a DsRed-tagged Golgi marker protein, the rat alpha-2,6-sialyltransferase. Therefore, we postulate that the majority of Arabidopsis cellulose synthase-like D proteins, under our experimental conditions, are likely located at the Golgi membranes. Furthermore, protease digestion of Golgi-rich vesicles revealed almost complete loss of reaction with the antibodies, even without detergent treatment of the Golgi vesicles. Therefore, the N-terminus of the Arabidopsis cellulose synthase-like D2 protein likely faces the cytosol. Combining this observation with the transmembrane domain predictions, we postulate that the large hydrophilic domain of this protein also faces the cytosol.     

The signal peptide of the rotavirus glycoprotein VP7 is essential for its retention in the ER as an integral membrane protein

The rotavirus glycoprotein VP7 has a cleavable signal peptide and is normally resident as an integral membrane protein in the ER of infected cells. A gene was constructed in which the VP7 H2 signal peptide was replaced by one from influenza hemagglutinin. COS cells transfected with this gene produced VP7 with the correct amino terminus, but the protein was rapidly secreted. Uncleaved VP7 from either precursor was not detected in cells after brief pulse-labeling, suggesting that the signal peptide was not acting as a temporary anchor; rather, it exerted its effect despite rapid cleavage. By splicing the H2 signal peptide onto another reporter protein, the malaria S-antigen, we demonstrated that H2 was necessary, but not itself sufficient, for targeting and retention. We propose that an interaction between the cleaved signal peptide and other downstream sequences in VP7 is required for retention of this protein in the ER as an integral membrane polypeptide.     

A subunit of mammalian signal peptidase is homologous to yeast SEC11 protein

Canine signal peptidase consists of a complex of five proteins (Evans, A. E., Gilmore, R., and Blobel, G. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 581-585). A cDNA encoding the 21-kDa subunit of the signal peptidase complex was isolated from a liver cDNA library using an 88-base pair probe, generated by the polymerase chain reaction. The 820-base pair cDNA was sequenced and found to encode a protein of 21,585 daltons. The deduced amino acid sequence from the canine cDNA was found to be 47% identical to the yeast SEC11 protein. SEC11 has been shown to be required for signal peptide cleavage, normal rate of secretion, and cell survival in Saccharomyces cerevisiae (B?hni, P. C., Deshaies, R. J., and Schekman, R. W. (1988) J. Cell Biol. 106, 1035-1042). It is, therefore, likely that the 21-kDa subunit of signal peptidase complex is the structural and functional homologue of the yeast SEC11 gene product.     

Influenza B virus BM2 protein is transported through the trans-Golgi network as an integral membrane protein

A bicistronic mRNA transcribed from the influenza B virus RNA segment 7 encodes two viral proteins, matrix protein M1 and uncharacterized small protein BM2. In the present study, we focused on the cytoplasmic transport and cellular membrane association of BM2. Immunofluorescence studies of virus-infected cells indicated that BM2 accumulated at the Golgi apparatus immediately after synthesis and then was transported to the plasma membrane through the trans-Golgi network. Localization of a set of BM2 deletion mutants revealed that the N-terminal half of BM2 (residues 2 to 50) was crucial for its transport; in particular, the deletion of residues 2 to 23, deduced to be a transmembrane domain, resulted in diffused distribution of the protein throughout the entire cell. Sucrose gradient flotation and biochemical analyses of the membrane showed that BM2 was tightly associated with cellular membranes as an integral membrane protein. Oligomerization of BM2 was demonstrated by coprecipitation of differentially epitope-tagged BM2 proteins. Taken together, these results strongly suggest that BM2 is integrated into the plasma membrane at the N-terminal hydrophobic domain as fourth membrane protein, in addition to hemagglutinin, neuraminidase, and NB, of the influenza B virus.     

Controlling the folding efficiency of an integral membrane protein

Research into the folding mechanisms of integral membrane proteins lags far behind that of water-soluble proteins, to the extent that the term protein folding is synonymous with water-soluble proteins. Hydrophobic membrane proteins, and particularly those with transmembrane alpha-helical motifs, are frequently considered too difficult to work with. We show that the stored curvature elastic stress of lipid bilayers can be used to guide the design of efficient folding systems for these integral membrane proteins. The curvature elastic stress of synthetic phosphatidylcholine/phosphatidylethanolamine lipid bilayers can be used to control both the rate of folding and the yield of folded protein. The use of a physical bilayer property generalises this approach beyond the particular chemistry of the lipids involved.     

SecY, a multispanning integral membrane protein, contains a potential leader peptidase cleavage site.

SecY is an Escherichia coli integral membrane protein required for efficient translocation of other proteins across the cytoplasmic membrane; it is embedded in this membrane by the 10 transmembrane segments. Among several SecY-alkaline phosphatase (PhoA) fusion proteins that we constructed previously, SecY-PhoA fusion 3-3, in which PhoA is fused to the third periplasmic region of SecY just after the fifth transmembrane segment, was found to be subject to rapid proteolytic processing in vivo. Both the SecY and PhoA products of this cleavage have been identified immunologically. In contrast, cleavage of SecY-PhoA 3-3 was barely observed in a lep mutant with a temperature-sensitive leader peptidase. The full-length fusion protein accumulated in this mutant was cleaved in vitro by the purified leader peptidase. A sequence Ala-202-Ile-Ala located near the proposed interface between transmembrane segment 5 and periplasmic domain 3 of SecY was found to be responsible for the recognition and cleavage by the leader peptidase, since a mutated fusion protein with Phe-Ile-Phe at this position was no longer cleaved even in the wild-type cells. These results indicate that SecY contains a potential leader peptidase cleavage site that undergoes cleavage if the PhoA sequence is attached carboxy terminally. Thus, transmembrane segment 5 of SecY can fulfill both of the two important functions of the signal peptide, translocation and cleavage, although the latter function is cryptic in the normal SecY protein.     

Regulation of signal peptidase by phospholipids in membrane: characterization of phospholipid bilayer incorporated Escherichia coli signal peptidase

Prokaryotic signal peptidases are membrane-bound enzymes. They cleave signal peptides from precursors of secretary proteins. To study the enzyme in its natural environment, which is phospholipid bilayers, we developed a method that allows us effectively to incorporate full-length Escherichia coli signal peptidase I into phospholipid vesicles. The membrane-bound signal peptidase showed high activity on a designed substrate. The autolysis site of the enzyme is separated from its catalytic site in vesicles by the lipid bilayer, resulting in a dramatic decrease of the autolysis rate. Phosphotidylethanolamine, which is the most abundant lipid in Escherichia coli inner membrane, is required to maintain activity of the membrane-incorporated signal peptidase. The maximal activity is achieved at about 55% phosphotidylethanolamine. Negatively charged lipids, which are also abundant in Escherichia coli inner membrane, enhances the activity of the enzyme too. Its mechanism, however, cannot be fully explained by its ability to increase the affinity of the substrate to the membrane. A reaction mechanism was developed based on the observation that cleavage only takes place when the enzyme and the substrate are bound to the same vesicle. Accordingly, a kinetic analysis is presented to explain some of the unique features of phospholipid vesicles incorporated signal peptidase, including the effect of lipid concentration and substrate-vesicle interaction.